Abstract

   Acute liver failure (ALF) represents a critical clinical syndrome
   marked by massive hepatocyte death and severe functional deterioration.
   While metabolic dysregulation is a recognized hallmark, the
   pathophysiological implications of iron metabolism disturbance in ALF
   progression remain poorly understood, which may unveil novel
   therapeutic targets. Using clinical samples and preclinical murine
   models, we identified ferroptosis as a predominant pathological feature
   in ALF-affected livers. Notably, pharmacological inhibition of
   ferroptosis significantly attenuated disease progression in
   experimental ALF. Mechanistically, dysregulation of the
   hepcidin-ferroportin (FPN) axis drives hepatic iron overload,
   precipitating ferroptotic cell death in ALF. The anti-rheumatoid
   arthritis drug auranofin restored hepcidin-FPN axis homeostasis and
   mitigated liver injury, though concomitant upregulation of
   proinflammatory cytokines limited its therapeutic potential.
   Strikingly, mesenchymal stromal cells (MSCs) demonstrated superior
   therapeutic efficacy, coordinately modulating the hepcidin-FPN axis
   while suppressing ferroptosis through PI3K/Akt/Nrf2 pathway activation.
   Our findings not only establish the causal relationship between
   hepcidin-FPN axis dysfunction and ferroptosis-driven liver injury, but
   also propose MSC-based therapy as a multifaceted strategy targeting
   both iron homeostasis and ferroptosis for ALF management.

   Keywords: Acute liver failure, Mesenchymal stromal cells, Ferroptosis,
   Iron metabolism, Hepcidin-ferroportin axis

Graphical abstract

   This study elucidates the pathophysiological mechanisms of
   hepcidin-ferroportin axis-mediated ferroptosis in the context of acute
   liver failure, as well as the role of mesenchymal stromal cells (MSCs)
   in modulating this axis to maintain iron homeostasis and inhibit
   ferroptosis, thereby mitigating liver failure. These findings indicate
   a promising therapeutic strategy for the management of acute liver
   failure.

   [37]Image 1
   [38]Open in a new tab

Abbreviations

   ALF
          acute liver failure

   TAA
          thioacetamide

   MSCs
          mesenchymal stromal cells

   FPN
          ferroportin

   TBI
          transferrin-bound iron

   TFR1
          transferrin receptor 1

   NTBI
          non-transferrin-bound iron

   SLC39A14
          solute carrier family 39 member 14

   IBA1
          Ionized Calcium-Binding Adapter Molecule 1

   HCs
          healthy controls

   PCA
          principle component analysis

   TEM
          transmission electron microscopy

   AST
          aspartate aminotransferase

   ALT
          alanine aminotransferase

   GSEA
          gene set enrichment analysis

   AUR
          Auranofin

   APAP
          acetaminophen

   Con A
          concanavalin A

   DEGs
          differentially expressed genes

   TFs
          transcription factors

   CCl4
          carbon tetrachloride

   Tfrc
          Transferrin receptor

   ROS
          reactive oxygen species

   GSH
          glutathione

   LPS
          lipopolysaccharide

   GalN
          d-galactosamine.

1. Introduction

   Acute liver failure (ALF) is a refractory, life-threatening liver
   disease characterized by the abrupt loss of a substantial number of
   hepatocytes and catastrophic impairment of liver function. As an
   end-stage condition, liver transplantation remains the sole definitive
   treatment for ALF [[39][1], [40][2]]. However, this approach is
   significantly constrained by the critical shortage of donated liver
   grafts. The short-term mortality rate of ALF exceeds one-third in the
   absence of timely liver transplantation [[41]3,[42]4]. ALF not only
   threatens patients' health and lives but also severely disrupts their
   occupational and social roles, placing immense pressure on both
   families and societies [[43]5]. Hence, it is imperative to elucidate
   the underlying mechanisms of ALF and develop innovative therapeutic
   strategies.

   The abrupt death of massive hepatocytes constitutes a pivotal event in
   triggering the progression of ALF. In recent years, various patterns of
   cell death have been identified [[44][6], [45][7], [46][8], [47][9],
   [48][10], [49][11]]. Ferroptosis, an iron-dependent and regulated form
   of cell death discovered in the past decade [[50][12], [51][13],
   [52][14]]. Previous researches have demonstrated that ferroptosis plays
   a key role in ALF, predominantly through mechanisms involving lipid
   peroxidation and oxidative stress [[53]8,[54]11,[55][15], [56][16],
   [57][17], [58][18]]. It should be noted that intracellular iron
   accumulation establishes the critical conditions necessary for lipid
   peroxidation and promotes its progression [[59]12,[60]14]. To fully
   illustrate the comprehensive landscape of ALF from the viewpoint of
   ferroptosis, it is crucial to delve deeper into the regulatory
   mechanisms of iron metabolism and explore potential intervention
   strategies that target these processes.

   The liver serves as the central regulator of iron metabolism and
   maintains systemic iron homeostasis through diverse mechanisms
   [[61]19,[62]20]. Imbalances of iron homeostasis can contribute to or
   exacerbate liver injury. Conversely, liver injury may also disrupt iron
   homeostasis, thereby establishing a vicious cycle [[63]21,[64]22].
   Therefore, understanding how to regulate iron metabolism and maintain
   its homeostasis is pivotal for preventing and treating ALF. Regarding
   iron uptake and transport, hepatocytes are capable of internalizing
   iron bound to transferrin (TBI) through the transferrin receptor 1
   (TFR1), as well as non-transferrin-bound iron (NTBI) via the solute
   carrier family 39 member 14 (SLC39A14) [[65]19,[66]20,[67]23]. It is
   NTBI, a highly toxic form of iron, that serves as a primary biomarker
   of iron overload and causes oxidative damage [[68]24]. Regarding iron
   storage and release, ferritin functions as the principal intracellular
   storage protein for iron, while ferroportin (FPN), the only known iron
   exporter in mammals, mediates its release when required. Notably,
   hepatocytes synthesize and secrete hepcidin, a key regulator of
   systemic iron homeostasis that precisely controls iron absorption and
   release. Specifically, hepcidin binds to ferroportin, triggering its
   degradation to regulate iron efflux and prevent toxic iron accumulation
   [[69]25,[70]26]. In alcoholic liver disease [[71]22], chronic viral
   hepatitis [[72]27] and liver fibrosis [[73]28], the abnormal expression
   of hepcidin or ferroportin impacts their pathophysiological processes.
   However, the precise role of the hepcidin-ferroportin axis in ALF and
   the potential intervention strategies remain to be elucidated.

   Mesenchymal stromal cells (MSCs), also known as medicinal signaling
   cells, are present in various adult tissues such as umbilical cord,
   bone marrow, and adipose tissue [[74]29]. MSCs have demonstrated
   considerable potential in treating liver diseases due to their
   remarkable paracrine effects, robust immunomodulatory properties, and
   tissue repair capabilities. [[75][30], [76][31], [77][32], [78][33],
   [79][34]]. Preclinical and preliminary clinical studies conducted by
   our group [[80]35,[81]36] and others [[82][37], [83][38], [84][39]]
   have demonstrated the efficacy and safety of MSCs, thereby highlighting
   their promising prospects for liver protection. However, the precise
   and comprehensive mechanisms underlying the therapeutic effects of MSCs
   in ALF remain incompletely elucidated, which somewhat hampers their
   extensive clinical application. In particular, the mechanism by which
   MSCs regulate iron metabolism remains poorly understood.

   Our present research elucidates the crucial role of ferroptosis in ALF,
   where dysregulation of the hepcidin-ferroportin signal axis disrupts
   iron homeostasis and initiates subsequent hepatic ferroptosis.
   Interestingly, we found that Auranofin (AUR), an FDA-approved drug for
   rheumatoid arthritis, exhibited hepatoprotective effects through the
   upregulation of hepcidin expression in the liver. These findings
   underscore the significant involvement of the hepcidin-ferroportin axis
   in ALF progression. However, AUR concurrently induced the expression of
   pro-inflammatory cytokines. Furthermore, our study highlights that MSCs
   modulate the hepcidin-ferroportin axis through the PI3K/Akt/Nrf2
   pathway, thereby finely regulating iron metabolism and effectively
   inhibiting ferroptosis, leading to enhanced therapeutic outcomes in
   ALF.

2. Materials and methods

2.1. Reagents

   Thioacetamide (#163678) and DMSO (#472301) were purchased from
   Sigma-Aldrich (St. Louis, MO, USA). Deferoxamine mesylate (#S5742),
   RSL3 (#S8155), ML385 (#S8790), MK-2206 (#S1078), and corn oil (#S6701)
   were obtained from Selleck Chemicals (Houston, TX, USA). Liproxstatin-1
   (#HY-12726), Auranofin (#HY-B1123), and PEG300 (#HY-Y0873) were
   purchased from MedChemExpress (MCE) (Shanghai, China). Tween 80
   (#MB2675) and normal saline (#MA0083) were purchased from MeilunBio
   (Dalian, China).

2.2. Human samples

   Our experiments with human liver samples were conducted in accordance
   with the research board protocol of the medical ethical committees of
   the Third Affiliated Hospital of Sun Yat-sen University. Informed
   consent was obtained from all subjects, and the experiment conforms to
   the principles set out in the WMA Declaration of Helsinki and the
   Department of Health and Human Services Belmont Report. Human liver
   tissues were acquired from nonviable grafts following liver
   transplantation procedures.

2.3. Animal experiments

   All animal procedures were approved by the Experimental Animal Care
   Commission of the Third Affiliated Hospital of Sun Yat-sen University.
   Six-to eight-week-old male C57BL/6 mice (control mice (n = 6), ALF mice
   (n = 20), MSCs treated mice (n = 12), Lip-1 treated mice (n = 12), DFOM
   treated mice (n = 12), ML385 treated mice (n = 12), Auranofin treated
   mice (n = 12)) were randomly divided into groups. Mice were raised in a
   specific pathogen-free (SPF) environment with constant temperature
   (22 ± 1 °C) and relative humidity (50 ± 5 %), and had free access to
   food and water. Next, mice were intraperitoneally injected with a
   single dose of TAA (150 mg/kg, dissolved in sterile saline) or saline.
   Twelve hours later, for MSCs administration, 2 × 10^6 cells (hUC-MSCs)
   were suspended in 0.2 ml of saline and transplanted into mice via tail
   vein injection. DFOM (100 mg/kg, dissolved in sterile saline), Lip-1
   (10 mg/kg, dissolved in 10 % DMSO + 40 % PEG300 + 5 % Tween-80 + 45 %
   sterile saline), ML385 (30 mg/kg, dissolved in 10 % DMSO + 40 %
   PEG300 + 5 % Tween-80 + 45 % sterile saline), and Auranofin (5 mg/kg,
   dissolved in 10 % DMSO + 90 % corn oil) were intraperitoneally injected
   into mice. The survival of all mice was recorded for 24 h, after which
   liver tissues were harvested for pathological analysis.

2.4. Serum biochemical Indexes detection

   Liver enzymes, including aspartate aminotransferase (AST) and alanine
   aminotransferase (ALT), were examined using the Hitachi Automatic
   Analyzer (3100; Hitachi Ltd., Tokyo, Japan). Cytokines and chemokines
   in serum were measured and quantified using the Milliplex® mouse
   immunology panel (Merck, MHSTCMAG-70K) according to the manufacturer's
   instructions. Each experimental group comprised three mice.

2.5. Measurements of non-heme iron, non-transferrin-bound iron (NTBI),
hepcidin and IL-6

   Tissue iron content was quantified using a colorimetric assay kit
   (I291, Dojindo Laboratories) following the manufacturer's protocol.
   Serum and tissue levels of non-transferrin-bound iron (NTBI) were
   measured using an ELISA kit (MM-46281M1, MEIMIAN) following the
   protocol provided by the manufacturer. Tissue hepcidin concentrations
   were measured with a Hepcidin ELISA kit (MM-44770M1, MEIMIAN) as per
   the manufacturer's guidelines. Serum interleukin-6 (IL-6) levels were
   assessed using a mouse IL-6 ELISA kit (550950, BD) in accordance with
   the manufacturer's specifications. Each experimental group comprised
   three mice.

2.6. Perls’ blue staining

   Tissue iron accumulation was analyzed using the Iron Stain Kit
   (Prussian Blue stain, Abcam, ab150674) in accordance with the
   manufacturer's protocol. Specifically, 5 μm paraffin-embedded liver
   sections underwent deparaffinization and dehydration. The sections were
   subsequently incubated in the iron stain solution for 10 min, followed
   by rinsing with distilled water and counterstaining with hematoxylin.
   The iron within the tissue sections reacted with acid ferrocyanide to
   produce a blue coloration. Microscopic images were acquired to document
   the staining results.

2.7. Histological analyses

   Liver tissue samples were subjected to formalin fixation and paraffin
   embedding. Five-micrometer-thick liver sections were stained with
   hematoxylin and eosin (H&E). The injured regions were analyzed using a
   NanoZoomer 2.0-RS whole-slide scanner (Hamamatsu Photonics, Hamamatsu,
   Japan). For immunohistochemical staining, the antibodies utilized are
   detailed in [85]Supplemental Table S1. Following heat-induced antigen
   retrieval and blocking with 5 % normal goat serum, the sections were
   incubated overnight at 4 °C with primary antibodies. After rewarming to
   room temperature and washing with PBS, the sections were incubated with
   biotin-labeled horse anti-mouse/rabbit IgG secondary antibodies for 1 h
   at room temperature. Finally, the sections were visualized using
   3,3′-diaminobenzidine (DAB) and counterstained with hematoxylin.

2.8. Immunofluorescence staining

   Liver tissue sections underwent deparaffinization, dehydration, and
   were treated with heat-induced antigen retrieval. They were then
   blocked with 5 % normal goat serum. The sections were incubated with
   primary antibodies overnight at 4 °C, then warmed to room temperature
   and rinsed with PBS. Subsequently, the sections were incubated with
   fluorescent-labeled secondary antibodies. Fluorescence microscopy
   images were acquired using an ECLIPSE Ni-E microscope (Nikon, Japan).

   AML12 cells were seeded on coverslips and exposed to RSL3 for 9 h,
   either as monocultures or cocultures with MSCs, or treated with ML385
   or MK-2206. Cells were fixed in 4 % paraformaldehyde for 30 min and
   permeabilized with 0.3 % Triton X-100 in 3 % BSA for 20 min at room
   temperature. Following this, cells were blocked with 3 % BSA for 1 h.
   Next, they were incubated with a primary antibody against p-Nrf2
   overnight at 4 °C. Cells were then incubated with the appropriate
   secondary antibody and counterstained with DAPI for 1 h at room
   temperature, with thorough washing between each step. Confocal
   laser-scanning microscopy images were captured using an LSM980
   microscope (Carl Zeiss, Germany).

2.9. Transmission electron microscopy

   Liver tissues were fixed in a solution containing 2.5 %
   paraformaldehyde and 2.5 % glutaraldehyde in Sorenson's phosphate
   buffer (pH 7.4). Subsequently, ultrathin sections (50–60 nm) were
   prepared and mounted on copper grids. The sections were subsequently
   stained with uranyl acetate and lead citrate to improve contrast,
   followed by examination using transmission electron microscopy (Hitachi
   H-7500).

2.10. RNA isolation and quantitative real time PCR

   Total RNA was extracted using TRIzol® Reagent (Vazyme) and subsequently
   reverse-transcribed into cDNA using HiScript II Q RT SuperMix for qPCR
   (+gDNA wiper) (Vazyme, Catalog No. R223-01). Quantitative real-time PCR
   (qRT-PCR) assays were conducted to evaluate the relative mRNA
   expression levels using AceQ® qPCR SYBR Green Master Mix (Vazyme). The
   mRNA levels of each gene were normalized to the internal control 18S
   rRNA. Primer sequences utilized in this study are detailed in
   [86]Supplemental Table S2.

2.11. Bulk RNA-seq analysis of liver tissues from ALF mouse models

   Total RNA was extracted from the liver tissues of ALF mouse models
   using TRIzol® Reagent (Vazyme), and the quality of the RNA was assessed
   by agarose gel electrophoresis. For RNA-seq, cDNA libraries were
   generated following the manufacturer's protocol with the MGIEasy RNA
   Library Prep Kit (catalog number 1000006384, MGI). Subsequently, gene
   expression profiling was performed on an Illumina HiSeq 2500
   instrument. The reads were mapped to the Ensembl Mus musculus (GRCm38.
   p4) reference genome using HISAT2 software. Differential gene
   expression (DEG) analysis was performed using DESeq2 to detect genes
   that were differentially expressed. Upregulated DEGs were used as input
   for Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment
   analysis and Gene Ontology (GO) analysis.

2.12. Immunoprecipitation and immunoblot analysis

   For the immunoprecipitation procedure, whole-cell lysates (n = 3) were
   prepared and incubated with the respective antibodies and Protein A/G
   beads (Pierce) overnight. The beads underwent six washes using a
   low-salt lysis buffer. Subsequently, the isolated protein was eluted
   with 1 × SDS loading buffer, fractionated by SDS-PAGE (12.5 %), and
   then blotted onto PVDF membranes. The membranes were then incubated
   with the appropriate antibodies (p-Nrf2, Akt, ferroportin, and Ub) and
   detected using an enhanced chemiluminescence (ECL) detection system.
   The antibodies used in this study are listed in [87]Supplemental Table
   S1. Protein expression levels were analyzed and quantified using ImageJ
   software.

2.13. Coculture assays

   After obtaining informed consent from the participants, the collection
   of human umbilical cords was performed. The hUC-MSCs were isolated and
   cultured as previously described [[88]40]. For coculture assays,
   3 × 10^5 AML12 cells were cocultured with MSCs using a Transwell
   culture system (0.4 μm pore size, Corning). 1 × 10^6 MSCs were seeded
   in the upper chamber. After treatment with RSL3 for 9 h, AML12 cells
   were cocultured with MSCs for 24 h.

2.14. Lipid peroxidation assay

   3 × 10^5 AML12 cells were seeded in 6-well plates and subsequently
   exposed to RSL3 for 9 h. Following this treatment, the cells were
   either maintained as monocultures, cocultured with MSCs, or treated
   with Liproxstatin-1 for 24 h. After this period, cells were washed with
   PBS and stained with C11-BODIPY (Thermo Fisher Scientific, D3861) at
   37 °C for 1 h. Cells were collected and subjected to centrifugation,
   followed by gentle resuspension in PBS. The fluorescence intensity of
   the cells was measured using a flow cytometer (BD Biosciences) and a
   fluorescence microscope (ECLIPSE Ni-E, Nikon, Japan). The data were
   analyzed using FlowJo software.

   Liver sections were incubated with C11-BODIPY, Dihydroethidium (DHE;
   Abcam, ab236206), or a combination of C11-BODIPY and PK Mito Deep Red
   (GENVIVO, PKMDR-1) at 37 °C in the dark for 1 h. Subsequently, the
   samples were washed three times with PBS and counterstained with DAPI
   (Beyotime, P0131) for nuclear visualization. Fluorescence intensity was
   quantified using a fluorescence microscope.

2.15. Statistical analysis

   Group differences were assessed using unpaired Student t-tests or
   one-way ANOVA followed by Bonferroni post hoc analyses (performed with
   GraphPad Prism 9.0, USA). For survival analysis, the Log-rank test was
   applied. Data are reported as means ± SD. Statistical significance was
   set at P-values <0.05, unless otherwise stated.

3. Results

3.1. Ferroptosis plays a pivotal role in driving the progression of acute
liver failure

   To elucidate the role of ferroptosis in the pathogenesis of acute liver
   failure (ALF), we employed multiple methods to assess the
   characteristic phenotypes in liver tissues from ALF patients, healthy
   controls (HCs), and an experimental ALF mouse model ([89]Fig. 1A).
   Analysis of ferroptosis markers demonstrated that GPX4 and FTH1 in
   liver tissues from ALF patients (n = 7) were significantly reduced
   compared to those from healthy controls (n = 7) ([90]Fig. 1B–E).
   Conversely, PTGS2 mRNA expression was markedly elevated in liver
   tissues from ALF patients ([91]Supplementary Fig. S1A). These findings
   indicate that ferroptosis might be a key factor in the development of
   ALF.

Fig. 1.

   [92]Fig. 1
   [93]Open in a new tab

   Ferroptosis is implicated in both patients and mouse models of ALF.(A)
   Workflow illustrating the procedural steps for phenotypic analysis in
   clinical samples, RNA-seq analysis of liver tissues from healthy
   controls (n = 3) and ALF mice (n = 3), and mechanistic investigations
   of potential hepatoprotective strategies. (B) Western blotting of GPX4
   and FTH1 protein expression in liver tissues from healthy controls
   (n = 7) and patients with acute liver failure (n = 7). Grayscale scan
   of GPX4 (C) and FTH1 (D) protein levels as depicted in (B). (E)
   Immunohistochemical analysis of GPX4 expression in liver tissues from
   healthy controls and patients with ALF. Scale bar: 50 μm. (F) Volcano
   plot showing the distribution of differentially expressed genes (DEGs)
   in liver samples from control mice versus those with acute liver
   failure. (G) GSEA demonstrating that the transcriptome of ALF is
   enriched in genes associated with cellular responses to oxidative
   stress. (H) Representative TEM images of the morphological
   characteristics of mitochondria in control and ALF liver tissues. Scale
   bar: 0.5 μm. (I) Representative pictures of livers from control and ALF
   mice treated with vehicle or ferroptosis inhibitors for 12 h. (J) Serum
   levels of AST and ALT in the control group, ALF + Saline group,
   ALF + Lip-1 group, and ALF + DFOM group (n = 3). All data are presented
   as means ± SD. ∗P < 0.05, ∗∗P < 0.01.

   To investigate how ferroptosis contributes to ALF, we established a
   TAA-induced mouse model and used RNA-seq to analyze liver tissue
   transcriptomes. A clear separation was observed between the ALF group
   and the control group in the PCA score plot ([94]Supplementary Fig.
   S1B). Gene expression analysis revealed high levels of ferroptotic
   drivers and markers in the ALF group, while ferroptotic suppressors
   were more prevalent in the control group ([95]Fig. 1F and
   [96]Supplementary Fig. S1C). Notably, the "cellular response to
   oxidative stress" category had the most upregulated genes in ALF liver
   tissues ([97]Fig. 1G). For the validation of ferroptosis phenotypes, in
   addition to the aforementioned alterations in the expression of
   specific genes, a critical aspect involves the characteristic
   morphological changes. Compared to the control group, transmission
   electron microscopy (TEM) revealed that the mitochondria in the liver
   tissue of the ALF group exhibited decreased size, higher membrane
   density, and a reduction or loss of cristae ([98]Fig. 1H). These
   characteristic alterations provide critical evidence for the occurrence
   of ferroptosis. In addition, to elucidate the distinct importance of
   ferroptosis in ALF liver tissue relative to other forms of regulated
   cell death, we examined the expression levels of genes associated with
   apoptosis and pyroptosis. Our results indicated that the expression
   levels of genes related to apoptosis and pyroptosis did not show
   significant changes in the ALF group ([99]Supplementary Fig. S1D).

   Furthermore, to elucidate the pivotal role of ferroptosis in the
   progression of ALF, we administered ferroptosis-specific inhibitors in
   mouse model to assess their effects. The results demonstrated that both
   Lip-1 and DFOM alleviated liver damage, as evidenced by substantial
   reductions in serum levels of aspartate aminotransferase (AST) and
   alanine aminotransferase (ALT) ([100]Fig. 1I, J). Notably, based on the
   results of HE staining of liver tissue ([101]Supplementary Fig. S1E),
   DFOM, which chelates iron, demonstrates a more pronounced effect in
   mitigating liver injury compared to Liprostatin-1, which reduces lipid
   peroxidation. This indicates that modulating iron metabolism and
   preventing iron overload may be a crucial therapeutic target for
   treating ALF. Collectively, these results have substantiated the
   critical significance of ferroptosis in the pathogenesis of ALF and
   illustrated that inhibiting ferroptosis could be a promising
   therapeutic strategy.

3.2. Iron overload, caused by a disorder of the hepcidin-ferroportin axis,
contributes to ferroptosis in ALF

   Ferroptosis is a kind of iron-dependent regulated cell death that is
   strongly associated with the disruption of iron metabolic homeostasis
   [[102][12], [103][13], [104][14], [105][15],[106]18]. Based on RNA-seq
   sequencing data from liver tissues of ALF model mice, gene set
   enrichment analysis (GSEA) indicated a markedly alteration in the
   expression of genes related to intracellular iron homeostasis in the
   ALF group. Conversely, the control group exhibited a significant
   enrichment of genes involved in iron-sulfur cluster binding ([107]Fig.
   2A and B), and these genes are of great significance in the regulation
   of iron metabolism. Furthermore, Perls's Blue Staining and tissue iron
   detection confirmed elevated iron levels in clinical ALF liver tissues
   ([108]Fig. 2C and [109]Supplementary Fig. S1F). Consistent findings
   were observed in the ALF mouse model, with increased liver iron levels,
   serum NTBI, and hepatic NTBI ([110]Fig. 2D–F). NTBI enters cells,
   leading to iron overload and subsequently inducing ferroptosis. These
   data indicate that an imbalance in iron metabolism in the liver leads
   to iron overload, which subsequently triggers ferroptosis in ALF.

Fig. 2.

   [111]Fig. 2
   [112]Open in a new tab

   Disruption of iron metabolism triggers ferroptosis in mice with ALF.
   (A) GSEA plot showing DEGs associated with intracellular iron
   homeostasis in comparison between the control and ALF groups. (B) GSEA
   plot showing DEGs associated with iron-sulfur cluster binding in
   comparison between control and ALF groups. (C) Perl's Prussian blue
   staining was performed on 5 μm paraffin-embedded liver sections
   obtained from healthy controls and patients with ALF. Scale bar: 50 μm.
   (D) Ferrous iron (Fe^2+) concentrations were measured in liver tissues
   from both the control and ALF groups (n = 3 for each group). (E) Liver
   tissue NTBI was quantified in both the control group and the ALF group,
   with three samples in each group. (F) Serum NTBI levels were quantified
   in control and ALF groups (n = 3). (G) Heatmaps illustrating the
   expression profiles of genes associated with iron metabolism. (H)
   Relative mRNA expression levels of Tfr1, Tfr2, Zip14, Pcbp1, Pcbp2,
   Ncoa4, Steap3, Cp, Ferroportin and Hepcidin in liver tissues from
   control and ALF groups (n = 3). (I) Immunoblot analysis of Ferroportin
   expression in liver tissues obtained from control and ALF groups
   (n = 3). (J) Quantification of Ferroportin protein expression in I. (K)
   Hepcidin levels in liver tissue were quantified in the control and ALF
   group (n = 3). (L) Immunohistochemical staining of HEPCIDIN in liver
   tissues from healthy controls and patients with ALF. Scale bar: 50 μm.
   All data are presented as means ± SD. ∗P < 0.05, ∗∗P < 0.01.

   Iron is a critical component that maintains a dynamic homeostasis
   through continuous uptake, transport, utilization, storage, and
   circulation within the body. This homeostasis is sustained by a complex
   and finely orchestrated process [[113]19,[114]20] ([115]Supplementary
   Fig. S1G). The core of the iron homeostasis regulatory mechanism is
   centered around the hepcidin-ferroportin (FPN) signaling axis,
   primarily relying on hepcidin secreted by hepatocytes
   [[116]19,[117]20]. Interestingly, prior research has documented that
   ALF patients display distinct alterations in iron parameters, such as
   hepcidin levels, which not only could serve as an independent predictor
   of disease outcome but also potentially play a causal role in the
   development of ALF [[118]21]. FPN, another crucial component in this
   signaling pathway, is the sole transporter known to facilitate iron
   export from cells. It is extensively expressed in macrophages and
   various other cell types. The interaction between ferroportin and
   hepcidin triggers the endocytosis and degradation of ferroportin, thus
   inhibiting iron efflux from cells [[119]25,[120]26,[121]41,[122]42]. In
   our study, RNA-seq data and qRT-PCR analysis revealed a significant
   upregulation of Slc39a14 (also known as Zip14), which is crucial for
   non-transferrin-bound iron (NTBI) uptake into cells, leading to iron
   overload ([123]Fig. 2G and H). Importantly, we observed a significant
   downregulation of hepcidin and FPN mRNA levels in the ALF group
   ([124]Fig. 2G and H). Protein level detection revealed a marked
   reduction in hepcidin expression and upregulation in ferroportin
   expression in ALF liver tissue ([125]Fig. 2I–K). Moreover, our
   subsequent research further demonstrated that in the liver tissues of
   ALF, ferroportin was predominantly overexpressed in macrophages rather
   than hepatocytes. In addition, immunostaining analyses further
   confirmed that hepcidin expression was significantly downregulated in
   liver tissues from ALF patients relative to healthy controls ([126]Fig.
   2L). These findings suggest that the dysfunction of the
   hepcidin-ferroportin signaling axis may be a critical factor
   contributing to iron overload in ALF.

3.3. Auranofin activates hepcidin-ferroportin axis to ameliorate ALF

   To clarify the role of the hepcidin-ferroportin regulatory axis in iron
   metabolism disorders and ferroptosis, we hypothesized that activating
   this axis might inhibit ferroptosis by regulating iron metabolism.
   Previous research by Professor Fudi Wang's group showed that Auranofin
   (AUR), an FDA-approved drug for rheumatoid arthritis, effectively
   promotes hepcidin mRNA expression [[127]43]. However, this study did
   not assess hepcidin protein levels in the liver tissues of AUR-treated
   mice. Therefore, we aimed to determine whether AUR could promote
   hepcidin protein expression and inhibit ferroptosis in ALF with fatal
   liver dysfunction. ALF mice that received 5 mg/kg body weight of AUR
   via intraperitoneal injection ([128]Fig. 3A) exhibited alleviation of
   liver injury, as evidenced by histological examination through HE
   staining and decreased serum AST and ALT levels ([129]Fig. 3B–E).
   Additionally, AUR improved the short term survival rate of ALF mice
   ([130]Supplementary Fig. S2A). Further analysis revealed a substantial
   downregulation of hepcidin expression in the ALF group, whereas AUR
   treatment led to a significant upregulation of hepcidin expression at
   both the mRNA and protein levels ([131]Fig. 3F and G). Consistent
   results were observed in the ELISA assay ([132]Fig. 3H).

Fig. 3.

   [133]Fig. 3
   [134]Open in a new tab

   Acute Auranofin administration enhances hepcidin expression and
   mitigates hepatic injury in mice with ALF. (A) Schematic illustration
   of the protocol for inducing ALF in mice and subsequent Auranofin
   administration. (B) Serum AST and ALT levels in the control group,
   ALF + Oil group, and ALF + Auranofin group (n = 3). (C) Representative
   gross specimens of livers from the control, ALF + Oil, and
   ALF + Auranofin groups. (D) Hematoxylin and eosin (H&E)-stained liver
   sections, visualized at magnifications of × 1.25 and × 20. Scale bar:
   500 μm, 50 μm. (E) The proportion of injured regions was quantified
   (n = 3). (F) mRNA expression levels of Hepcidin in liver tissues from
   control, ALF + Oil, and ALF + Auranofin groups (n = 3). (G)
   Immunohistochemical staining of hepcidin was performed on liver tissues
   obtained from three groups: control, ALF treated with oil, and ALF
   treated with auranofin. Scale bar: 50 μm. (H) Hepcidin levels in liver
   tissue were quantified in the control group, the ALF + Oil group, and
   the ALF + Auranofin group (n = 3). (I) Immunoblot analysis of
   Ferroportin expression in liver tissues from control, ALF + Oil, and
   ALF + Auranofin groups (n = 3). (J) Quantification of Ferroportin
   expression in I. (K) Immunofluorescent staining of IBA1 and Ferroportin
   was performed on liver sections derived from the control, ALF + Oil,
   and ALF + Auranofin groups. Scale bar: 50 μm. All data are presented as
   means ± SD. ∗P < 0.05, ∗∗P < 0.01.

   It has been reported that ferroportin is primarily regulated at the
   post-translational level by hepcidin. Specifically, hepcidin binds to
   ferroportin, triggering its internalization and then degradation,
   thereby restricting the amount of iron released from cells
   [[135]25,[136]26]. Therefore, we further investigated the protein
   expression and polyubiquitination of ferroportin (FPN) in liver
   tissues. Our results demonstrated that the protein level of ferroportin
   (FPN) was significantly elevated in the ALF group, while treatment with
   AUR markedly reduced ferroportin protein expression and concurrently
   enhanced its polyubiquitination ([137]Fig. 3I, J and [138]Supplementary
   Fig. S2E). It is widely recognized that macrophages play a crucial role
   in restoring and sequestering iron by engulfing damaged red blood cells
   (RBCs). When needed, they release iron via the iron export protein
   ferroportin, thus maintaining iron balance. This process is
   meticulously controlled by the hepcidin-ferroportin axis. In our
   investigation, we noted a rapid increase in IBA-positive macrophages
   (MoMFs) surrounding dead hepatocytes following ALF, consistent with the
   observations made by Professor Gao's group [[139]44]. Importantly, we
   discovered that ferroportin expression was predominantly localized to
   these macrophages rather than hepatocytes, and AUR treatment markedly
   decreased ferroportin levels in macrophages ([140]Fig. 3K and
   [141]Supplementary Fig. S2D). These findings indicate that modulating
   the hepcidin-ferroportin regulatory axis could represent a potential
   therapeutic approach for the management of ALF. So, is Auranofin the
   optimal choice for treating ALF?

   Previous studies have shown that short-term administration of AUR not
   only increases the expression of hepcidin mRNA in the liver but also
   that of interleukin-6 (IL-6) [[142]43]. Consequently, we analyzed the
   expression levels of inflammatory cytokines and chemokines. The
   findings revealed that, compared with untreated ALF mice, the mRNA
   levels of Il6 and Cxcl1 were significantly elevated in the AUR-treated
   group, whereas the expression of Ccl2 was reduced, and there was no
   notable impact on Cxcl10 expression ([143]Supplementary Fig. S2B).
   Furthermore, ELISA confirmed a substantial increase in the serum IL-6
   protein level in ALF mice following AUR treatment ([144]Supplementary
   Fig. S2C). However, during the progression of ALF, disorder of the
   immune system is particularly pronounced, and the inflammatory response
   plays a critical role in disease progression [[145]2]. A significant
   amount of inflammatory factors, such as IL-6, released after hepatocyte
   death can trigger excessive immune activation, leading to systemic
   inflammatory response syndrome (SIRS) [[146]45]. These inflammatory
   responses are not confined to the liver but also affect other organ
   systems, contributing to multiple organ dysfunction, which is a major
   factor in the high mortality rate associated with ALF [[147]2,[148]46].
   Given these findings, the elevated IL-6 expression induced by AUR may
   not be beneficial for improving overall prognosis and liver function in
   ALF patients, particularly in the long term. Therefore, we pose the
   question: is there a more suitable treatment option than Auranofin?
   This question merits further investigation to develop more effective
   treatment strategies for ALF patients.

3.4. MSCs activates hepcidin-ferroportin axis to maintain iron homeostasis

   It is widely acknowledged that iron homeostasis in the body represents
   a dynamic equilibrium process. Key regulators of iron metabolism, such
   as hepcidin and ferroportin, must function within an optimal range to
   maintain iron balance. Both excessively high and low levels of hepcidin
   can disrupt iron homeostasis, thereby adversely impacting overall
   health. A clinical study on Rusfertide (a hepcidin mimetic) has shown
   that hepcidin agonists may be associated with a series of side effects,
   some of which are potentially life-threatening [[149]47]. Therefore,
   the cornerstone of treating iron-related diseases lies in restoring and
   maintaining iron homeostasis rather than merely modulating specific
   components. A mild yet potent modulator, rather than a simple agonist
   or analogue of hepcidin, would more effectively restore iron
   homeostasis in ALF. Mesenchymal stromal cells (MSCs), a type of adult
   stem cell with multi-target and multi-functional properties, exhibit
   potent secretory and proliferative capabilities. They protect the liver
   by mechanisms including the inhibition of inflammatory responses,
   modulation of immune function, suppression of apoptosis, and promotion
   of cell proliferation [[150][32], [151][33], [152][34]]. These features
   position MSCs as a highly promising candidate for achieving this
   therapeutic objective. Our previous studies have shown that
   transplantation of mesenchymal stromal cells (hUC-MSCs) derived from
   human umbilical cord can effectively alleviate acute liver injury
   caused by acetaminophen (APAP) [[153]48] or concanavalin A (Con A)
   [[154]36,[155]49]. Not surprisingly, our results demonstrated that MSCs
   infusion significantly inhibited the activation of the IL-17 signaling
   pathway in TAA-induced ALF mice ([156]Supplementary Fig. S3A).
   Additionally, we measured cytokine expression levels and found that
   serum levels of IL-6 and MCP-1 were elevated significantly, which could
   be mitigated by MSCs infusion ([157]Supplementary Fig. S3B and C).
   These findings indicate that MSCs can compensate the abovementioned
   disadvantage of Auranofin, specifically the unexpectedly elevated
   inflammatory factors. However, the precise mechanism by which MSCs
   repair liver injury remains not fully elucidated, particularly from the
   perspective of iron metabolism. Based on the aforementioned
   experimental results, we hypothesized that MSCs might alleviate ALF by
   regulating the hepcidin-ferroptotin signaling axis and maintaining iron
   metabolism homeostasis. Firstly, our findings demonstrated that MSCs
   treatment significantly increased the mRNA levels of hepcidin and
   ferroportin relative to those in the ALF group ([158]Fig. 4A). Then we
   found that MSCs infusion led to a significant rescue in hepcidin
   expression at protein level ([159]Fig. 4B–D). Conversely, the protein
   level of ferroportin was significantly elevated in the ALF group,
   whereas MSCs treatment markedly downregulated ferroportin protein
   expression ([160]Fig. 4E, F and [161]Supplementary Fig. S3D). Given
   that ferroportin protein levels are regulated by hepcidin at the
   post-translational level, we further examined whether the reduction in
   ferroportin protein in MSCs-treated ALF mice was due to enhanced
   polyubiquitination. Not coincidentally, our findings indicated that
   MSCs treatment increased the polyubiquitination of ferroportin
   ([162]Fig. 4G). These data suggest that MSCs treatment modulates the
   hepcidin-ferroportin axis. Next, we investigated whether activation of
   the hepcidin-ferroportin axis by MSCs could maintain iron homeostasis.
   As anticipated, we observed a remarkable reduction in serum and hepatic
   non-transferrin-bound iron (NTBI) levels in the group receiving MSCs
   infusion ([163]Fig. 4H–J). Collectively, our data indicate that MSCs
   treatment significantly alleviates iron overload in TAA-induced ALF, at
   least in part, through the hepcidin-ferroportin axis.

Fig. 4.

   [164]Fig. 4
   [165]Open in a new tab

   MSCs regulate the hepcidin-ferroportin axis to maintain iron
   homeostasis. (A) mRNA levels of Hepcidin and Ferroportin in liver
   tissues from control, ALF + Saline and ALF + MSCs groups (n = 3). (B)
   Representative immunofluorescence staining was conducted to assess
   hepcidin expression in hepatocytes from three groups: control,
   ALF + Saline, and ALF + MSCs. Scale bar: 20 μm. (C) Immunohistochemical
   staining of hepcidin in liver tissues from three groups: control, ALF
   treated with saline, and ALF treated with MSCs. Scale bar: 20 μm. (D)
   Hepcidin concentrations in liver tissue were quantified using ELISA in
   three groups: the control group, the ALF + Saline group, and the
   ALF + MSCs group (n = 3). (E) Immunoblot analysis of Ferroportin
   expression in liver tissues from control, ALF + Saline, and ALF + MSCs
   groups (n = 3). (F) Quantification of Ferroportin expression via
   greyscale scanning in (E). (G) The liver tissue lysates were subjected
   to immunoprecipitation (IP) with anti-Ferroportin or IgG as a negative
   control and immunoblot analysis with anti-ubiquitin (n = 3). (H)
   Ferrous iron (Fe^2+) concentrations were detected in liver tissues of
   the control, ALF + Saline and ALF + MSCs groups (n = 3). (I) Liver
   tissue NTBI was measured in control, ALF + Saline and ALF + MSCs groups
   (n = 3). (J) Serum NTBI was measured in control, ALF + Saline and
   ALF + MSCs groups (n = 3). All data are presented as means ± SD.
   ∗P < 0.05, ∗∗P < 0.01.

3.5. MSCs regulate hepcidin expression in a PI3K/Akt/Nrf2-dependent manner

   To investigate the molecular mechanisms through which MSCs influence
   the hepcidin-ferroportin axis, we analyzed the gene expression profile
   of liver tissue from mice with ALF. A total of 969 upregulated
   differentially expressed genes (DEGs) were identified in the MSCs
   treatment group compared to the ALF group. To further assess the
   potential core transcription factors (TFs) involved in the regulation
   of the hepcidin-ferroportin axis in the MSCs treatment group, we
   inputted the 969 upregulated DEGs, along with ferroptosis-related
   drivers, markers, and suppressors, into the TRRUST database. Overlap
   analysis revealed that 39 overlapping TFs were obtained, including
   Nrf2, Tp53, Nfκb, Sp1, Stat3, Ep300, Jun, Egr1, and others ([166]Fig.
   5A and [167]Supplemental Table S3 and S4). Consistent with these
   findings, qRT-PCR analyses showed that MSCs treatment restored the
   expression of Nrf2, Tp53, Sp1, and Stat3 ([168]Fig. 5B). Of note, Nrf2,
   a well-known regulator of antioxidant responses and iron metabolism,
   has been reported to stimulate hepcidin expression upon activation,
   thereby maintaining iron homeostasis, which may serve as a therapeutic
   target for iron-related diseases [[169][50], [170][51], [171][52]].
   Based on these findings, we hypothesize that there might be a
   connection between Nrf2 and hepcidin following MSCs treatment. To
   investigate this, we analyzed the phosphorylation levels of Nrf2
   (p-Nrf2) in liver tissues from each group and found that MSCs treatment
   significantly elevated the levels of p-Nrf2 compared to those in the
   ALF group ([172]Fig. 5C and D). Coincidentally, AUR treatment resulted
   in a significant increase in the levels of p-Nrf2 ([173]Supplementary
   Fig. S2F and G). Additionally, gene expression profiling demonstrated
   that several Nrf2 downstream genes were upregulated following MSCs
   treatment ([174]Supplementary Fig. S5A). Next, we aimed to identify the
   upstream regulators of p-Nrf2. KEGG enrichment analysis revealed that
   the Rap1-PI3K-Akt signaling pathway was markedly enhanced following
   MSCs treatment ([175]Supplementary Fig. S5B). At the protein level,
   immunohistochemistry analysis confirmed that MSCs treatment
   significantly upregulated the expression of phosphorylated Akt
   ([176]Supplementary Fig. S5C). Furthermore, to investigate the
   potential regulatory relationship between Nrf2 and Akt, we performed
   co-IP and immunoblot assays. The results demonstrated that the
   association between p-Nrf2 and Akt was significantly enhanced following
   MSCs treatment ([177]Fig. 5E).

Fig. 5.

   [178]Fig. 5
   [179]Open in a new tab

   PI3K/Akt/Nrf2 signaling pathway mediates hepcidin expression induced by
   MSCs treatment. (A) Venn diagram illustrating the intersection of
   transcription factors (TFs) in upregulated DEGs from MSCs-treated ALF
   mice and genes associated with ferroptosis. (B) qRT-PCR was employed to
   validate the expression levels of several candidate transcription
   factors (TFs) in liver tissues obtained from the control, ALF + Saline
   and ALF + MSCs groups (n = 3). (C) Immunoblot analysis of
   phosphorylated Nrf2 (p-Nrf2) and total Nrf2 in liver tissues obtained
   from control, ALF + Saline, and ALF + MSCs groups (n = 3). (D)
   Quantification of p-Nrf2 expression via greyscale scanning in (C). (E)
   Liver tissue lysates were subjected to immunoprecipitation (IP) using
   anti-Akt antibodies, with IgG serving as a negative control, followed
   by immunoblot analysis with anti-p-Nrf2 antibodies. (F) AML12 cells
   were treated with RSL3 for 9 h, followed by mono-culturing or
   co-culturing with MSCs for an additional 24 h. Subsequently, immunoblot
   analysis was conducted to examine the levels of p-Nrf2, Nrf2, p-Akt,
   and Akt. (G) Quantification of p-Nrf2 and p-Akt expression via
   greyscale scanning is presented in (F). (H) Microscopy was performed on
   AML12 cells that were treated with RSL3 and either monocultured or
   cocultured with MSCs for 24 h, followed by immunostaining of
   phosphorylated Nrf2 (p-Nrf2) using a specific primary antibody (stained
   green). Nuclei were counterstained with DAPI (blue). Scale bar: 20 μm.
   (I) Immunohistochemical analysis of p-Nrf2 expression in liver tissues
   from control, ALF + Saline, and ALF + MSCs groups. Scale bar: 20 μm.
   All data are presented as means ± SD. ∗P < 0.05, ∗∗P < 0.01.

   To further substantiate these findings, we developed an in vitro
   coculture system wherein AML12 cells were exposed to either DMSO or
   RSL3, followed by monoculture or coculture with MSCs, with parallel
   groups receiving Liproxstatin-1 (Lip-1) treatment for 24 h
   ([180]Supplementary Fig. S4A). AML12 cells treated with RSL3, a known
   ferroptosis inducer, exhibited a notable increase in the characteristic
   "ballooning" morphology associated with ferroptotic cell death. Both
   MSCs coculture and Lip-1 treatment significantly mitigated RSL3-induced
   ferroptosis, as evidenced by a marked reduction in the number of cells
   undergoing ferroptosis ([181]Supplementary Fig. S4B and C).
   Furthermore, the expression of Gpx4 protein was found to be elevated in
   AML12 cells either cocultured with MSCs or treated with Lip-1
   ([182]Supplementary Fig. S4D and E). Consistent with these findings,
   RSL3 administration resulted in a significant increase in lipid-ROS
   levels, which were comparably attenuated by MSCs coculture or Lip-1
   treatment ([183]Supplementary Fig. S4F–H). Collectively, these data
   indicate that MSCs exert an inhibitory effect on ferroptosis in vitro.

   Next, in order to investigate whether MSCs inhibit ferroptosis by
   modulating the PI3K/Akt/Nrf2 signaling pathway, we performed in vitro
   experiments and demonstrated that treatment with the ferroptosis
   inducer RSL3 led to a significant reduction of p-Nrf2 accumulation in
   the nucleus. However, MSCs not only upregulated the phosphorylation
   levels of Nrf2 (p-Nrf2) and Akt (p-Akt) ([184]Fig. 5F and G), but also
   facilitated the translocation of p-Nrf2 into the nucleus ([185]Fig. 5H
   and I). To further confirm whether there is a direct causal
   relationship between Nrf2 signaling and hepcidin expression mediated by
   MSCs, we evaluated the recovery of liver injury and hepcidin expression
   in ALF mice treated with MSCs with or without the presence of ML385, a
   well-tolerated Nrf2 inhibitor ([186]Fig. 6A). As expected, treatment
   with ML385 counteracted the therapeutic effects of MSCs in ALF mice, as
   evidenced by serum levels of AST, ALT, and HE staining ([187]Fig.
   6B–D). Consistently, ML385 treatment downregulated hepcidin expression
   in ALF mice treated with MSCs ([188]Fig. 6E and F). Additionally,
   inhibition of Akt and Nrf2 reduced p-Nrf2 nuclear accumulation in
   AML12 cells co-cultured with MSCs ([189]Supplementary Fig. S6A). All
   together, these findings suggest that MSCs promote hepcidin expression
   through a PI3K/Akt/Nrf2-dependent mechanism.

Fig. 6.

   [190]Fig. 6
   [191]Open in a new tab

   Inhibition of Nrf2 impairs the recovery from liver injury and hampers
   hepcidin expression mediated by MSCs.(A) Schematic diagram of ALF mice
   treated with vehicle, MSCs or Nrf2 inhibitors (ML385). (B) Serum AST
   and ALT levels were measured in the control group, ALF + Saline group,
   ALF + MSCs group, and ALF + MSCs + ML385 group (n = 3). (C) Hematoxylin
   and eosin (H&E)-stained liver sections examined at × 1.25 and × 20
   magnification. Scale bar: 500 μm, 50 μm. (D) The percentage of injured
   areas was quantified in three independent samples (n = 3). (E)
   Immunohistochemical staining of hepcidin in liver tissues obtained from
   control, ALF + Saline, ALF + MSCs, and ALF + MSCs + ML385 groups. Scale
   bar: 50 μm. (F) mRNA levels of hepcidin in liver tissues from control,
   ALF + Saline, ALF + MSCs, and ALF + MSCs + ML385 groups (n = 3). All
   data are presented as means ± SD. ∗P < 0.05, ∗∗P < 0.01. #P < 0.05,
   ##P < 0.01.

3.6. MSCs inhibits ferroptosis to alleviate ALF

   Based on the above findings, we have established that the disruption of
   the hepcidin-ferroportin axis plays a critical role in ferroptosis in
   ALF mice. Moreover, mesenchymal stromal cells (MSCs) transplantation
   activates hepcidin, thereby maintaining iron homeostasis through the
   regulation of the PI3K/Akt/Nrf2 signaling pathway. To further explore
   whether MSCs treatment can inhibit ferroptosis and alleviate ALF, we
   conducted additional experiments. As anticipated, mice treated with
   MSCs exhibited reduced liver injury compared to the control group, as
   evidenced by lower serum AST and ALT levels ([192]Fig. 7A). Consistent
   with this, treatment with MSCs markedly enhanced the survival rate of
   mice with ALF ([193]Fig. 7B). We then assessed the expression of
   ferroptosis markers in liver tissues from each group. The results
   demonstrated that the mRNA levels of Gpx4 were significantly
   downregulated, while those of Hmox1, Ptgs2, and Slc7a11 were
   upregulated in ALF mice. However, MSCs treatment markedly restored Gpx4
   expression and reduced the expression of Hmox1, Ptgs2, and Slc7a11
   ([194]Fig. 7C). Similar trends were also observed at the protein level.
   Following MSCs infusion, the protein levels of Transferrin receptor
   (Tfrc), Gpx4, and Fth1 were significantly increased compared to those
   in the ALF group ([195]Fig. 7D and E). Immunostaining further validated
   these observations ([196]Fig. 7F and [197]Supplementary Fig. S7A). To
   evaluate oxidative stress in liver tissues, we conducted C11-BODIPY and
   DHE staining for ROS detection. The results demonstrated that ROS
   production was significantly higher in the ALF group than in the
   control group, while MSCs treatment effectively attenuated ROS
   production ([198]Supplementary Fig. S7B). Moreover, subcellular
   localization analysis demonstrated that the elevated ROS predominantly
   accumulated in the mitochondria, and MSCs infusion significantly
   mitigated mitochondrial lipid peroxidation ([199]Fig. 7G).
   Additionally, transmission electron microscopy (TEM) revealed that
   mitochondria in the MSC-treated group showed signs of normalization
   compared to the ALF group ([200]Fig. 7H). Collectively, these results
   indicate that MSCs treatment effectively mitigates ALF by inhibiting
   ferroptosis.

Fig. 7.

   [201]Fig. 7
   [202]Open in a new tab

   Treatment with MSCs inhibits ferroptosis, thereby alleviating ALF.(A)
   Serum AST and ALT levels in the control group, the ALF + Saline group,
   and the ALF + MSCs group (n = 3). (B) Survival curves of mice in the
   ALF + Saline and ALF + MSCs groups. (C) mRNA levels of Gpx4, Hmox1,
   Ptgs2 and Slc7a11 in liver tissues from control, ALF + Saline and
   ALF + MSCs groups (n = 3). (D) Immunoblot analysis of Tfrc, Gpx4 and
   Fth1 protein expression in liver tissues from control, ALF + Saline,
   and ALF + MSCs groups (n = 3). (E) Quantification of Tfrc, Gpx4 and
   Fth1 expression levels was performed through greyscale scanning as
   shown in (D). (F) Immunochemical analysis of Gpx4 expression in liver
   tissues from control, ALF + Saline, and ALF + MSCs groups. Scale bar:
   50 μm. (G) Liver tissues were probed with C11-BODIPY and PK-mito. Scale
   bar: 50 μm. (H) Representative TEM images illustrating the
   mitochondrial morphology in the control, ALF + Saline, and ALF + MSCs
   groups. Scale bar: 0.5 μm. All data are presented as means ± SD.
   ∗P < 0.05, ∗∗P < 0.01.

4. Discussion

   Acute liver failure is a grave condition characterized by abrupt
   occurrence, rapid progression, and high fatality rates
   [[203]1,[204]2,[205]10]. It can lead to various metabolic dysfunctions,
   such as disruptions in iron metabolism, thereby exacerbating the
   progression of ALF. In the current study, our findings indicate that
   ferroptosis is a critical pathophysiological mechanism in the
   development of ALF. Mechanistically, the dysregulation of
   hepcidin-ferroportin axis results in iron overload, thus triggering
   ferroptosis in the liver. Interestingly, we identified a novel
   therapeutic application of Auranofin, an anti-rheumatic drug, which
   exerts hepatoprotective effects by upregulating hepcidin. More
   importantly, our results demonstrated that MSCs rectify iron metabolism
   disorders and inhibit ferroptosis by activating hepcidin-ferroportin
   axis, and effectively alleviate ALF.

   Previous studies have suggested that apoptosis and pyroptosis are the
   predominant forms of cell death in ALF [[206][7], [207][8], [208][9],
   [209][11]]. Even earlier studies support oncotic necrosis as the
   primary mode of cell death, though some overlap with apoptosis,
   pyroptosis and ferroptosis signaling events fuels ongoing debates
   [[210]6]. Although inhibitors targeting these modes of cell death may
   offer protective effects in experimental models, no specific
   pharmacological inhibitors have been applied clinically to date.
   Therefore, further research is essential to elucidate additional and
   comprehensive mechanisms underlying ALF pathogenesis. Notably, the
   liver exhibits a high susceptibility to ferroptosis. First, the liver's
   critical role in iron storage and regulation predisposes it to abnormal
   iron metabolism and accumulation, creating an environment conducive to
   ferroptosis [[211]19]. Second, the liver's rich lipid content serves as
   a substrate for lipid peroxidation and influences the activity of lipid
   metabolic enzymes involved in ferroptosis. Additionally, the liver's
   active metabolism and detoxification functions expose it to oxidative
   stress, which further promotes ferroptosis [[212]15,[213]18].
   Consequently, ferroptosis is closely linked to the pathogenesis and
   progression of various liver diseases. Our studies, along with those of
   other research groups, have demonstrated the potential of ferroptosis
   inhibitors in mitigating liver injury. Targeted modulation of
   ferroptosis is anticipated to emerge as a novel therapeutic strategy
   for liver diseases.

   Some studies have shown that disorders of iron metabolism play an
   important role in the pathogenesis and progression of certain liver
   diseases, such as hereditary hemochromatosis (HH) and non-alcoholic
   fatty liver disease [[214]14,[215]15,[216]18]. In the context of acute
   chemical or drug-induced liver injury/liver failure, existing
   literature mainly investigates the process of ferroptosis from the
   perspective of oxidative damage and impairment to antioxidant systems.
   For example, in carbon tetrachloride (CCl4)-induced acute liver injury,
   GPX4 and SLC7A11 are inhibited, impairing the liver's ability to
   scavenge reactive oxygen species (ROS), resulting in lipid peroxidation
   and a large number of liver cells undergoing ferroptosis [[217]53].
   Additionally, acetaminophen (APAP) overdose, representing as the most
   common cause of ALF in many countries, leads to decreased levels of
   glutathione (GSH) within hepatocytes, accompanied by increased lipid
   peroxidation [[218]6]. Similar phenomena have also been observed in
   mouse models of ALF induced by lipopolysaccharide (LPS) and
   d-galactosamine (GalN) [[219]8]. Based on these findings, some scholars
   have proposed the development of antioxidant therapies as a promising
   treatment strategy for ALF. However, others have challenged this
   notion, arguing that antioxidants alone are insufficient to effectively
   mitigate the systemic inflammatory response syndrome (SIRS) associated
   with ALF [[220][45], [221][46]]. Nonetheless, disorders of iron
   metabolism, a critical biochemical process driving ferroptosis in ALF,
   have not received comparable attention. A clinical study revealed
   significant alterations in iron metabolism parameters among patients
   with ALF, primarily characterized by reduced serum hepcidin levels, an
   independent predictor of ALF outcomes [[222]21]. Consistent with these
   findings, our study also identified decreased hepcidin levels in ALF
   liver tissue. Furthermore, we demonstrated that upregulating hepcidin
   expression via Auranofin administration or MSCs transfusion exerts
   beneficial effects on alleviating ALF. However, an unexpected increase
   in inflammatory factors, such as IL-6, was observed following Auranofin
   treatment. While IL-6 is known to initiate hepatocyte proliferation
   through JAK-STAT3, PI3K/Akt, and YAP/Notch pathways to promote cell
   cycle entry and DNA synthesis, thereby playing a potential role in
   liver regeneration, it cannot be overlooked that during acute liver
   injury, elevated IL-6 levels mediate inflammation by activating
   non-parenchymal cells such as macrophages, T cells, and hepatic
   stellate cells via NF-κB and STAT3 pathways, potentially exacerbating
   hepatic damage at high concentrations or with persistent elevation
   [[223]54]. Given that Auranofin exhibits the adverse effect of
   upregulating the expression of inflammatory factors, it is imperative
   to reassess its clinical application and develop optimal alternatives
   for the treatment of ALF. It is noteworthy that despite the observation
   of excess iron accumulation in patients with acute liver failure, this
   does not necessarily indicate that reducing iron levels would be
   beneficial. Considering iron's essential role in human health,
   therapeutic strategies for acute liver failure should focus on
   restoring or maintaining iron homeostasis. Given that iron homeostasis
   is regulated by the coordinated action of multiple key molecules and
   signaling pathways, achieving this goal may prove challenging with
   compounds or drugs targeting only a single or fixed target. Compared to
   traditional single-target or fixed-target small molecule compounds,
   multi-target therapeutic strategies such as MSCs therapy may provide
   unparalleled advantages in the treatment of complex clinical syndromes
   like ALF.

   MSCs, referred to as mesenchymal stromal cells or medicinal signaling
   cells, exhibit significant potential in regenerative medicine
   [[224]55]. This potential is attributed to their unique
   characteristics, such as diverse origins, minimal ethical concerns, low
   immunogenicity, and robust secretory functions that surpass their
   self-renewal and differentiation capabilities. Importantly, MSCs
   possess chemotactic migration properties, the ability to mitigate
   oxidative damage, suppress inflammation, promote cell proliferation,
   and exert immunomodulatory effects, positioning them as a promising
   therapeutic option for liver diseases [[225]29,[226][31], [227][32],
   [228][33], [229][34]]. Based on the antioxidant properties of MSCs,
   Feiyan Lin et al. demonstrated that these cells maintain SLC7A11
   stability, thereby activating system XC^- and ultimately inhibiting
   ferroptosis [[230]53]. However, the role of MSCs in modulating iron
   metabolism for the treatment of ALF has been underexplored in previous
   studies. Notably, our findings indicate that MSCs regulate iron
   metabolism via the hepcidin-ferroportin axis to inhibit ferroptosis,
   offering a novel perspective distinct from prior research. Nonetheless,
   further investigation is required to fully elucidate the effector
   molecules involved in MSC-mediated regulation of iron
   metabolism-related signaling pathways.

   The findings from our research group [[231]35,[232]36,[233]48,[234]49]
   and other investigators [[235]31,[236]37,[237]53] have provided
   substantial evidence supporting the remarkable hepatoprotective role of
   MSCs. Their safety and efficacy have been validated through extensive
   clinical trials [[238]35,[239][37], [240][38], [241][39]].
   Consequently, MSCs have demonstrated considerable therapeutic potential
   and broad application prospects in the treatment of ALF (29, 31–33).
   Notably, on December 18, 2024, the FDA granted approval for an
   MSC-based therapy to treat steroid-resistant acute graft-versus-host
   disease (SR-aGVHD), marking the world's first allogeneic
   "off-the-shelf" cell therapy approved by the FDA. This represents a
   significant milestone in the clinical application of MSCs therapy.
   Beyond the FDA-approved indications for GVHD, further in-depth
   fundamental research is essential to fully explore the potential of
   MSCs therapy in treating ALF. Precise indication selection and the
   design of appropriate intervention strategies are critical for the
   success of MSCs clinical research. Specifically, several key factors
   must be considered: the source of MSCs, the optimal therapeutic time
   window for MSCs administration, the dosage per transfusion, the
   frequency and intervals between transfusions, and whether MSCs should
   undergo pretreatment or genetic modification to enhance efficacy.
   Additionally, elucidating the comprehensive mechanisms of MSCs therapy
   in ALF remains equally important. With the ongoing advancement of MSCs
   clinical trials, it is anticipated that an increasing number of
   MSC-based therapeutic products will be developed and applied clinically
   in the future, offering new hope for patients suffering from conditions
   such as ALF.

   In conclusion, to the best of our knowledge, the present study is the
   first to identify the hepcidin-ferroportin axis as a promising
   therapeutic target for ALF based on extensive functional experiments.
   Dysregulation of the hepcidin-ferroportin axis and the resulting iron
   metabolism disorder contribute to ferroptosis, thereby providing a more
   comprehensive understanding of ALF pathophysiology. Furthermore,
   Auranofin, an FDA-approved anti-rheumatic drug, has exhibited novel
   hepatoprotective properties by regulating iron metabolism and
   inhibiting ferroptosis through upregulation of hepcidin expression.
   More importantly, we have uncovered a novel mechanism by which MSCs can
   effectively regulate iron metabolism by activating the
   hepcidin-ferroportin axis via the PI3K/Akt/Nrf2 pathway. These findings
   may have substantial implications for the treatment strategies of ALF.

CRediT authorship contribution statement

   Jinyong He: Writing – original draft, Visualization, Methodology,
   Investigation, Data curation. Cong Du: Writing – review & editing,
   Writing – original draft, Supervision, Investigation, Funding
   acquisition, Conceptualization. Cuiping Li: Software, Investigation.
   Wei Li: Software, Investigation. Jinlan Qiu: Software, Investigation.
   Mingpeng Ma: Software, Investigation. Yunhao Chen: Software,
   Methodology. Qi Zhang: Writing – review & editing, Supervision, Project
   administration, Funding acquisition, Conceptualization.

Financial support and sponsorship

   This work is supported by grants from the National Nature Science
   Foundation of China (82170674, 82425010, 82300749, 82000620), Guangdong
   Basic and Applied Basic Research Foundation (2023A1515010583) and China
   Postdoctoral Science Foundation (2022M723610, 2024T171076).

Declaration of competing interest

   The authors declare that they have no competing interests.

Footnotes

   ^Appendix B

   Supplementary data to this article can be found online at
   [242]https://doi.org/10.1016/j.redox.2025.103657.

Contributor Information

   Cong Du, Email: ducong3@mail.sysu.edu.cn.

   Qi Zhang, Email: zhangq27@mail.sysu.edu.cn.

Appendix B. Supplementary data

   The following are the Supplementary data to this article.
   Multimedia component 1
   [243]mmc1.docx^ (18.4KB, docx)
   Multimedia component 2
   [244]mmc2.docx^ (37.7KB, docx)

figs1.

   [245]figs1
   [246]Open in a new tab

figs2.

   [247]figs2
   [248]Open in a new tab

figs3.

   [249]figs3
   [250]Open in a new tab

figs4.

   [251]figs4
   [252]Open in a new tab

figs5.

   [253]figs5
   [254]Open in a new tab

figs6.

   [255]figs6
   [256]Open in a new tab

figs7.

   [257]figs7
   [258]Open in a new tab

Data availability

   Data will be made available on request.

References